RESEARCH ARTICLE Open Access

Icariin alleviates osteoarthritis by inhibitingNLRP3-mediated pyroptosisYan Zu1* , Yue Mu2, Qiang Li1, Shu-Ting Zhang1 and Hong-Juan Yan1

Abstract

Background: Osteoarthritis (OA) is the common chronic degenerative joint bone disease that is mainly featured byjoint stiffness and cartilage degradation. Icariin (ICA), an extract from Epimedium, has been preliminarily proven toshow anti-osteoporotic and anti-inflammatory effects in OA. However, the underlying mechanisms of ICA onchondrocytes need to be elucidated.

Methods: LPS-treated chondrocytes and monosodium iodoacetate (MIA)-treated Wistar rats were used as modelsof OA in vitro and in vivo, respectively. LDH and MTT assays were performed to detect cytotoxicity and cell viability.The expression levels of NLRP3, IL-1β, IL-18, MMP-1, MMP-13, and collagen II were detected by qRT-PCR andWestern blotting. The release levels of IL-1β and IL-18 were detected by ELISA assay. Caspase-1 activity wasassessed by flow cytometry. Immunofluorescence and immunohistochemistry were used to examine the level ofNLRP3 in chondrocytes and rat cartilage, respectively. The progression of OA was monitored with hematoxylin-eosin (H&E) staining and safranin O/fast green staining.

Results: ICA could suppress LPS-induced inflammation and reduction of collagen formation in chondrocytes.Furthermore, ICA could inhibit NLRP3 inflammasome-mediated caspase-1 signaling pathway to alleviate pyroptosisinduced by LPS. Overexpression of NLRP3 reversed the above changes caused by ICA. It was further confirmed inthe rat OA model that ICA alleviated OA by inhibiting NLRP3-mediated pyroptosis.

Conclusion: ICA inhibited OA via repressing NLRP3/caspase-1 signaling-mediated pyroptosis in models of OA invitro and in vivo, suggesting that ICA might be a promising compound in the treatment of OA.

Keywords: Icariin, Osteoarthritis, NLRP3 inflammasome, Caspase-1 signaling, Pyroptosis

IntroductionOsteoarthritis (OA) is a long-term joint bone disorderwhich is recognized by the progressive damage of jointstructure, such as articular cartilage and subchondralbone [1]. OA is now considered as a significant clinicalproblem worldwide as the increase of aging populations.Currently, the treatment agents have failed to block theprogression of OA, thus safer and better-tolerated drugsfor OA therapy are needed [2]. During the progressionof OA, pathological events occur in the cartilage includ-ing oxidative stress, inflammation, apoptosis, loss of car-tilage matrix, and autophagy. The studies of geneticmouse models demonstrated that growth factors, such

as transforming growth factor-β (TGF-β), Wnt3a, andIndian hedgehog, as well as signaling molecules, such asβ-catenin, Smad3, and HIF-2α, are involved in OA pro-gression [3–5]. Inflammation and extracellular matrix(ECM) loss are increasingly recognized as essentialdrivers of OA cartilage injury [6, 7]. Understanding theregulation of these changes may provide effective ther-apy for OA.Pyroptosis is a caspase-1-required inflammasome-

mediated programmed cell death signal pathway. Pyrop-tosis can lead to IL-1β and IL-18 production as well asmacrophage cell lysis [8]. Distinct from apoptosis andsimple cell necrosis, the process of pyroptosis is proin-flammatory and mediated by nod-like receptor protein-3(NLRP3) inflammasome and caspase-1 signaling. A re-cent study demonstrated that NLRP inflammasomesNLRP1 and NLRP3 participated in the mediation of

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

* Correspondence: zuyan@ncut.edu.cn1School of Mechanical and Material Engineering, North China University ofTechnology, No.5, Jinyuanzhuang Road, Shijingshan District, Beijing 100144,People’s Republic of ChinaFull list of author information is available at the end of the article

Zu et al. Journal of Orthopaedic Surgery and Research (2019) 14:307 https://doi.org/10.1186/s13018-019-1307-6

lipopolysaccharide (LPS)-induced fibroblast-like synovio-cytes (FLSs) pyroptosis [9]. The inhibition of these twoNLRP inflammasomes led to a significant decline ofpyroptosis-related cytokines, suggesting NLRP1 andNLRP3 inflammasomes might play an important role inthe pathogenesis of OA. One study presented that hy-droxyapatite crystals of a particular size and shape wereable to promote the secretion of proinflammatory cyto-kines IL-1β and IL-18 from murine macrophages, whichwas dependent of NLRP3 inflammasome [10]. And thelevel of NLRP3, IL-1β, and IL-18 is increased in the syn-ovial membrane of knee osteoarthritis rats, indicatingthe involvement of inflammasome in OA [11]. However,the role and underlying mechanism of NLRP3-mediatedpyroptosis in OA is still unknown.The Chinese traditional medical herb Epimedium had

long been used to attenuate the OA process. Epimediumcould significantly moderate OA condition and declinethe expression level of urokinase plasminogen activator(uPA), uPA receptor (uPAR), and urokinase plasminogenactivator inhibitor (PAI) in rabbit OA model [12]. Icariin(ICA) which is the extract of Epimedium was identifiedas the effective constituent, showing anti-osteoporoticand anti-inflammatory effects [12]. Studies revealed thatICA prevented OA inflammation and chondrocytesapoptosis though activation of autophagy via inhibitingNF-κB signaling pathway [13]. Furthermore, ICA exerteda chondroprotective effect through the inhibition ofMMP-1, MMP-3, and MMP-13 or the suppression ofosteoprotegerin (OPG), receptor activator of nuclear fac-tor kappa-B ligand (RANKL), and receptor activator ofnuclear factor kappa-B (RANK) system via MAPK path-way in IL-1β-stimulated chondrosarcoma cells [14, 15].However, the molecular mechanisms of ICA alleviatingOA and its relationship with NLRP3 inflammasome arenot fully understood.In this study, we demonstrated that ICA reduced LPS-

induced pyroptosis and the inhibition of collagen forma-tion through suppressing NLRP3 inflammasome-mediatedcaspase-1 signaling pathway. The effect of ICA alleviatingOA by inhibiting NLRP3-mediated pyroptosis was furtherconfirmed in rat OA model. Thus, ICA may play thera-peutic roles in OA treatment.

Materials and methodsIsolation and culture of rat chondrocytesCartilage slices were shaved from the articular surface ofboth knee joints from adult male Wistar rats. Pronasesolution (2%, 1.5 mL) was used to treat the pooled cartil-age slices for 20 min at 37 °C, and the rest of pure cartil-age was rinsed with phosphate-buffered saline (PBS) fortwice. Collagenase solution (1.5 mL) was then added andthe samples were incubated again at 37 °C until theextracellular matrix was completely digested. The

chondrocytes were then filtered through a mesh screenand the resulted single-cell suspension was centrifugedat 1500g for 10 min. The cells were transferred to T-25cell culture flasks and incubated with complete Dulbec-co’s modified Eagle’s medium (DMEM) at 37 °C in a 5%CO2 incubator.

Cell transfectionThe full-length NLRP3 sequence was constructed intopcDNA3.1, and these reconstructed plasmids werenamed pcDNA3.1-NLRP3. The empty plasmid wasemployed as the negative control. Cells were inoculatedin a 6-well plate at a density of 3 × 105/well. When cellconfluence reached 80%, cells were transfected with dif-ferent plasmids using Lipofectamine 2000 kit (InvitrogenInc., Carlsbad, CA, USA). The target plasmids (4 μg) andlipofectamine 2000 (10 μL) were separately diluted with250 μL serum-free Opti-MEM (Gibco Company, NY,USA). The two dilutions were allowed to stand at roomtemperature for 5 min and mixed evenly. The mixturewas added to the culture well after being placed for 20min, and then cultured in an incubator with 5% CO2 at37 °C. The complete medium was employed for subcul-ture after 6 h, and cells were collected 48 h afterward.

Cell viabilityRat chondrocytes were seeded in 96-well plates (1 × 104

cells/well), and these cells were treated with ICA (1, 2.5,5, 10, 20 and 40 μM) for 24 h. Then, 20 μL of MTT (5mg/mL in PBS) was added to each well, and the plateswere incubated at 37 °C for 4 h. After removing the su-pernatants, 150 μL dimethylsulfoxide was supplementedto each well. The plates were then shaken for 10 min,and optical density (OD) values were detected and re-corded at 570 nm using a microplate reader (Bio-Rad,CA, USA).

Flow cytometry analysis of caspase-1 activityTo measure pyroptosis in chondrocytes, the level of ac-tive caspase-1 was detected by caspase-1 Detection Kit(ImmunoChemistry, Bloomington, MN, USA). FAM-YVAD-FMK was used to label active caspase-1 enzymeand determined by flow cytometry following the manu-facturer’s instruction. Briefly, the cell suspension was in-cubated with FAM-YVAD-FMK for 60 min at 37 °C inthe dark. After centrifugation, the supernatant was re-moved by aspiration and the cell pellet was washed twicewith 1 × wash buffer. Cells were resuspended in PI stain-ing buffer and kept on ice. The cells were analyzed by aflow cytometer (BD Biosciences, San Jose, CA, USA),and pyroptosis was defined as double positive for FAM-YVAD-FMK and PI staining.

Zu et al. Journal of Orthopaedic Surgery and Research (2019) 14:307 Page 2 of 12

LDH release and ELISA assayCytotoxicity was measured by quantifying lactate dehydro-genase (LDH) release in the cell supernatant by colorimet-ric assay following the manufacturer’s instructions(Clontech, Mountain View, CA, USA). The levels of IL-1βand IL-18 in chondrocyte supernatants were measuredusing IL-1β or IL-18 enzyme-linked immunosorbent assay(ELISA) kits (R&D Systems, USA) according to the manu-facturer’s instructions. The optical density (OD) of the ex-tracted lysate supernatant was measured and recorded at630 nm with a microplate reader.

ImmunofluorescenceThe chondrocytes, cultured on cover slips for 48 h, werefixed with 4 % paraformaldehyde solution (Santa CruzBiotechnology, USA), followed by treatment with blockingsolution containing 5 % donkey serum and 0.1 % Triton(Sigma Aldrich, Germany) in PBS. Subsequently, the sam-ples were incubated with the polyclonal anti-Rat NLRP3antibody (1:60, abcam, USA) overnight at 4 °C followed bywashing three times with PBS and incubating with theanti-mouse IgG secondary antibody (1:200, abcam, USA)and DAPI (Roche Applied Science, Germany) for 1 h atroom temperature. The images were collected by anOlympus FluoView FV10i confocal microscope.

OA rat modelMale Wistar rats were purchased from human SJA La-boratory Animal Co., Ltd (Changsha, China). All proce-dures were performed according to the internationallyaccepted ethical guidelines. Rats were housed understandard laboratory conditions with free access to foodand water with room temperature at 22 °C and humidityabout 45~55%. After 1 week of acclimatization, rats weredivided randomly into three groups (n = 5 per group) asfollows: (1) control group: no monosodium iodoacetate(MIA) injection; (2) OA group: MIA injection; and (3)OA + ICA group (MIA + ICA): ICA injection 2 weekspost MIA injection. According to the grouping, rats weredirectly injected with MIA (40 mg/mL in 0.9% NaCl so-lution, 50 μL) or 0.9% NaCl solution (50 μL) into theintra-articular space of the right knee except for the con-trol group. And after 2 weeks, rats in the OA + ICAgroup were injected with ICA (20 μM, 0.3 mL). After 32days, cartilage block with subchondral bone was isolatedfor the subsequent experiments.

Hematoxylin-eosin (H&E) staining and safranin O/fastgreen stainingCartilage block with subchondral bone was cut into 1.0cm × 1.0 cm × 0.5 cm-sized blocks and these blocks werefixed for 3 days with neutral formalin solution, followedby decalcification for 14 days in 30% formic acid solutionand dehydration with ethanol in conventionally gradient.

The samples were embedded in paraffin and cut intoslices at 5 μm. For hematoxylin-eosin (H&E) staining,the cartilage samples were stained with Harris alumhematoxylin (Fuzhou Maixin Biotechnology, China) for5 min after dewaxing and hydration. Following washingfor 10 s in 0.5% hydrochloric acid alcohol, the slices werestained in eosin (Fuzhou Maixin Biotechnology, China)for 40 s. After dehydration, transparency, and mountingwith neutral balsam, the samples were visualized under amicroscope. The nucleus of the chondrocytes appearedblue and the other tissues appeared pink.For safranin O/fast green staining, the samples were

stained with Weigert’s iron hematoxylin and then rinsedwith water before placed in 0.2% fast green solution(Shanghai Sangon Biological Engineering Technology,China) for 1 min, 1% ethylic acid solution for 30 s, and0.1% safranin O solution (Shanghai Sangon BiologicalEngineering Technology, China) for 15 min. The sampleswere then dehydrated, transparentized, and mountedwith neutral balsam. The normal cartilage appeared redand the background appeared green.

Immunohistochemistry (IHC)Paraffin wax-embedded tissue sections of subchondralbone were used for IHC. The sections were dewaxed inxylene and dehydrated with gradient ethanol. After anti-gen retrieval in sodium citrate buffer, the sections wereincubated with anti-rat NLRP3 antibody (Abcam, USA)overnight at 4 °C or at 37 °C. After three washes in PBS(3 min each time), the tissues were incubated with bio-tinylated IgG (1:200, Solarbio, China) at 37 °C for 30min, followed by 3 times PBS rinse. Afterward, fresh di-aminobenzidine (DAB) (Boster Biological Technology,China) was added for coloration for 1~2min to visualizethe antibody-antigen complexes. The slices with brownor yellow cytoplasm were considered to be positive. Fivetypical fields were randomly selected from each slice forobservation and counted under an optical microscope(Nikon, Tokyo, Japan).

Western blottingThe cells were lysed using ice-cold RIPA buffer contain-ing protease inhibitor (Boster Biological Technology,China). Proteins from cartilage samples were extractedas previous reports [16, 17]. Briefly, the frozen tissueswere mechanically pulverized, sonicated, and stored incold extraction buffer (Roche) at 4 °C overnight. Thesupernatant after centrifugation was collected and pro-teoglycans (PGs) was cleared by cetylpyridinium chloride(CPC). After another centrifugation, the supernatant wastreated with methanol and following centrifugation, thepellet was resuspended in 2-DE sample buffer (GEHealthcare Life Sciences, Sweden). The proteins werequantified using the bicinchoninic acid protein assay kit

Zu et al. Journal of Orthopaedic Surgery and Research (2019) 14:307 Page 3 of 12

(BCA kit) according to the manufacturer’s instructions.An equal amount of proteins (20 μg) were separated by10% SDS-PAGE and transferred to polyvinylidene fluor-ide (PVDF) membranes (Millipore, USA), which werethen blocked in 5% bull serum albumin (BSA, Sigma,USA) in TBST for 2 h at room temperature. The mem-branes were then incubated with primary antibody at4 °C overnight. The primary antibodies were as follows:NLRP3 (1:2000, abcam, USA), ASC (1:2000, abcam,USA), caspase-1 (1:2000, abcam, USA), GSDMD (1:5000,abcam, USA), IL-1β (1:5000, abcam, USA), IL-18 (1:5000, abcam, USA), MMP-1(1:2000, abcam, USA),MMP-13 (1:2000, abcam, USA), collagen II (1:2000,abcam, USA), and GAPDH (1:1000, abcam, USA). Afterwashing three times with TBST, the membranes wereincubated with HRP-conjugated corresponding second-ary antibodies (Goat Anti-Rabbit IgG, 1:5000, abcam,USA) for 1 h at room temperature. After washing againwith TBST for three times, the immunoreactive proteinswere visualized with the ECL western detection kit(Thermo Fisher Scientific, USA). ImageJ software wasused to quantify the density of each band.

RNA extraction and quantitative real time-PCR (qRT-PCR)Total RNA from chondrocytes and cartilage tissue wasextracted using Trizol reagent (Thermo Fisher Scientific,USA) and reverse-transcribed to cDNA with RT MasterMix (Takara Japan). The RT-PCR was performed withSYBR Master Mix using StepOne-Plus system (ABI,USA) by denaturing at 95 °C for 30 s, annealing at 60 °Cfor 1 min and extending at 95 °C for 5 s. The primer se-quences were as follows: NLRP3 forward (5'-GTAGGTGTGGAAGCAGGACT-3') and reverse (5'-CTTGCTGACTGAGGACCTGA-3′), IL-1β forward (5′-CAGCAGCATCTCGACAAGAG-3′) and reverse (5′-CATCATCCCACGAGTCACAG-3′), IL-18 forward (5′-ATGCCTGATATCGACCGAAC-3′) and reverse (5′-TGGCACACGTTTCTGAAAGA-3′), and GAPDH for-ward (5′-CAAGTTCAACGGCACAG-3nc) and reverse(5′-CCAGTAGACTC CACGACAT-3′). The gene ex-pression was analyzed by 2−ΔΔCt method using GAPDHas the internal control.

Statistical analysisAll experiments were conducted at least three times anddata from the representative experiment was shown. Datawere present as mean ± standard deviation (SD). GraphpadPrism 5.0 (GraphPad Software Inc., USA) was used for stat-istical analysis and statistical evaluation was performed byStudent’s t test (two-tailed) between two groups or one-wayanalysis of variance (ANOVA) followed by Tukey post hoctest for multiple comparison. p values<0.05 were consideredstatistically significant.

ResultsLPS induces inflammation and reduction of collagenformation in chondrocytesThe expressions of NRLP3, IL-1β, and IL-18 in chondrocyteswere examined by qRT-PCR following treatment of LPS. Asshown in Fig. 1a, the expression of NRLP3, IL-1β, and IL-18was dramatically induced by LPS in rat chondrocytes. Inaddition, the result of the ELISA assay also indicated in-creased release level of IL-1β and IL-18 (Fig. 1b). The expres-sion levels of MMP-1, MMP-13, collagen II, NRLP3, IL-1β,and IL-18 in LPS-treated rat chondrocytes were monitoredby Western blotting (Fig. 1c, d). The accumulation of MMP-1, MMP-13, NRLP3, IL-1β, and IL-18 proteins were inducedafter incubation with LPS for 12 h, while the expression ofcollagen II shows an opposite trend. In combination, theseresults indicated that LPS induces inflammation of chondro-cytes and decreases collagen formation.

ICA suppresses LPS-mediated chondrocytes injury andpyroptosisMTT assays were employed to analyze the effects of ICAon rat chondrocytes viability. ICA decreased cell viabilityin a concentration-dependent manner, and the applica-tion of ICA over 40 μM was able to dramatically inhibitchondrocytes viability (Fig. 2a). The lactate dehydrogen-ase (LDH) leakage was monitored as an indicator of pyr-optosis. LPS incubation increased the extracellular LDH,while the application of ICA rescued the cells from theleakage of LDH (Fig. 2b). Moreover, the reverse effect ofICA on LDH was dose-dependent (Fig. 2b). The effectsof ICA on pyroptosis in rat chondrocytes were furtheranalyzed by detecting the expression of IL-1β and IL-18.As shown in Fig. 2c, d, LPS-induced IL-1β and IL-18 ex-pression was reversed by ICA, and this suppressionfunction of ICA was dose-dependent. Consistent withthe qRT-PCR result, the release level of IL-1β and IL-18protein was also decreased upon the application of ICA(Fig. 2e, f), and the suppression of IL-1β and IL-18 re-lease level was shown to be ICA concentration-dependent. The expression of collagen formation andpyroptosis-related molecules, including MMP-1,MMP13, collagen II, IL-1β, and IL-18, were analyzed byWestern blotting. Data showed that ICA dose-dependently suppressed LPS-induced promotion ofMMP-1, MMP13, IL-1β, and IL-18, while preventedLPS-reduced collagen II expression (Fig. 2g, h). Takentogether, these results suggested that ICA may suppressLPS-induced chondrocytes injury and pyroptosis.

ICA inhibits LPS-induced activation of NLRP3inflammasome and pyroptosis-related caspase-1 signalingpathwayTo further investigate the effect of ICA on LPS regulatedNLRP3 expression, immunofluorescence microscopy

Zu et al. Journal of Orthopaedic Surgery and Research (2019) 14:307 Page 4 of 12

was performed. The results demonstrated that LPS in-creased the expression of NLRP3 in rat chondrocytes(Fig. 3a). It was also found ICA disrupted LPS in-creased NLRP3 expression dose-dependently. Mean-while, the chondrocytes pyroptosis was evaluated byflow cytometry by measuring the activity of caspase-1.As shown in Fig. 3b, c, the activity of caspase-1 wasincreased by LPS in chondrocytes when comparedwith the control group. However, the application ofICA suppressed caspase-1 activity in a dose-dependent manner (Fig. 3b, c). The levels of NLRP3,ASC, caspase-1, and GSDMD were evaluated byWestern blotting, and the results revealed that ICAnegatively regulated these pyroptosis-related proteinsexpression induced by LPS (Fig. 3d, e). These resultsindicated that ICA can reverse LPS-activated NLRP3and caspase-1 signaling pathway, thus suppress LPS-induced pyroptosis in chondrocytes.

ICA suppresses pyroptosis of rat chondrocytes throughinhibitingNLRP3 signalingTo investigate the regulatory mechanism of ICA in ratchondrocytes pyroptosis, NLRP3 overexpression wasconstructed. As shown in Additional file 1: Figure S1,the green fluorescence intensity of the transfection effi-ciency increased gradually over time after transfection ofthe NLRP3 overexpression vector. And the green fluor-escence intensity was the strongest after transfection for48 h; thus, the transfection time was 48 h for subsequentexperiments. The expression of NLRP3 was verified byqRT-PCR and Western blotting and dramatically in-creased NLPR3 mRNA and protein level was detected inpcDNA3.1-NLRP3 transfected chondrocytes, suggestingthat NLPR3 was successfully overexpressed (Fig. 4a, b).The LDH leakage, IL-1β, and IL-18 level in chondrocyteswere monitored. Consistent with former results, theLDH level and the expression levels of IL-1β and IL-18

Fig. 1 Effects of LPS on chondrocytes inflammation and collagen formation. a The mRNA levels of NRLP3, IL-1β, and IL-18 were analyzedby qRT-PCR in rat chondrocytes treated with LPS. b The release levels of IL-1β and IL-18 in the cell supernatant of LPS treated ratchondrocytes was checked by ELISA assay. c The expression level of collagen formation and pyroptosis-related protein was analyzed byWestern blotting in rat chondrocytes treated with LPS. d Quantitative analysis of protein band gray in c. GAPDH was used as internalcontrol. All the results were shown as mean ± SD (n = 3), which were three separate experiments performed in triplicate. *p < 0.05and **p < 0.01

Zu et al. Journal of Orthopaedic Surgery and Research (2019) 14:307 Page 5 of 12

Fig. 2 Effects of ICA on chondrocytes injury and pyroptosis. a The cell activity was detected by MTT assay in chondrocytes treated with variousconcentrations of ICA for 24 h. b Leakage of LDH was assessed by ELISA assay in chondrocytes treated with LPS or ICA. The mRNA levels of IL-1β(c) and IL-18 (d) were analyzed by qRT-PCR in chondrocytes treated with LPS or ICA. The release level of IL-1β (e) and IL-18 (f) were analyzed byELISA assay in chondrocytes treated with LPS or ICA. g The protein level of collagen formation and pyroptosis-related protein was analyzed byWestern blotting in rat chondrocytes treated with LPS or ICA. h Quantitative analysis of protein band gray in g. GAPDH was used as internalcontrol. All the results were shown as mean ± SD (n = 3), which were three separate experiments performed in triplicate. *p < 0.05 and **p < 0.01

Zu et al. Journal of Orthopaedic Surgery and Research (2019) 14:307 Page 6 of 12

were induced by LPS, while the application of ICA sup-pressed this change (Fig. 4c–e). However, overexpressionof NLRP3 could reverse the effects of ICA on LDH, IL-1β, and IL-18 release levels (Fig. 4c–e). The levels of IL-1β, IL-18, NLRP3, ASC, caspase-1, and GSDMD wereevaluated by Western blotting, and the results confirmedthat overexpression of NLRP3 attenuated the protectioneffects of ICA on chondrocyte pyroptosis (Fig. 4f, g).

Taken together, these results suggested that ICA-mediated suppression of chondrocytes pyroptosis isexerted by regulating NLRP3.

ICA alleviates rat OA by inhibiting NLRP3To investigate the potential protective effects of ICA onosteoarthritis in vivo, we constructed a ratOA model byinjecting MIA. The results of H&E and safranin O/fast

Fig. 3 ICA exerted negative effects on LPS-induced NLRP3 expression and pyroptosis. a Immunofluorescent was used to detect NLRP3 inchondrocytes treated with LPS or ICA. b Flow cytometry analysis was used to measure caspase-1 activity in chondrocytes treated with LPS or ICA.c Quantitative statistics of doublepositive for caspase-1 and PI stain. d The protein level of pyroptosis-related protein was analyzed by Westernblotting in chondrocytes treated with LPS or ICA. e Quantitative analysis of protein band gray in d. GAPDH was used as internal control. All theresults were shown as mean ± SD (n = 3), which were three separate experiments performed in triplicate. *p < 0.05 and **p < 0.01

Zu et al. Journal of Orthopaedic Surgery and Research (2019) 14:307 Page 7 of 12

green staining showed that in rat OA model, there was asevere total erosion of cartilage compared to the controlgroup (Fig. 5a). In contrast, ICA-treated rats showed lesssevere destructions, which was revealed by a reduced

loss of safranin O staining and surfaced regularity(Fig. 5a). The expression level of NLRP3 was detectedusing IHC staining. As shown in Fig. 5a, b, the rat OAmodel exhibited a significant increase in the number of

Fig. 4 ICA suppressed chondrocytes injury and pyroptosis through NLRP3. a The mRNA level of NLRP3 was validated by qRT-PCR in chondrocytes. b Theprotein level was assessed by Western blotting in chondrocytes. GAPDH was used as internal control. c Leakage of LDH was assessed by ELISA assay inchondrocytes. The release level of IL-1β (d) and IL-18 (e) were analyzed by ELISA assay in chondrocytes. f The protein level of pyroptosis-related proteinwas analyzed by Western blotting in rat chondrocytes. g Quantitative analysis of protein band gray in e.GAPDH was used as internal control. All the resultswere shown as mean ± SD (n= 3), which were three separate experiments performed in triplicate. *p< 0.05 and **p< 0.01

Zu et al. Journal of Orthopaedic Surgery and Research (2019) 14:307 Page 8 of 12

NLRP3-positive cells compared to the control group,while the application of ICA dramatically reduced posi-tive cell number in the rat cartilage tissue. These resultsclearly demonstrated that ICA is able to modulate theprogression of OA and suppress NLRP3.

ICA attenuates NLRP3-mediated pyroptosis and increasescollagen formation in rat OA modelsTo clarify the mechanism of ICA on osteoarthritis pro-gress, qRT-PCR was used to determine the expression ofpyroptosis-related OA marker genes in rat OA models.The results showed that ICA significantly reduced theexpression of NLRP3, IL-1β, and IL-18 in OA rats(Fig. 6a–c), which were consistent with the results of invitro chondrocytes experiments. To further validate theabove results, we also assessed the effects of ICA on theaccumulation of pyroptosis-related OA marker proteinsby Western blotting. The results revealed that the pro-tein level of MMP-1, MMP-13, NRLP3, IL-1β, and IL-18were suppressed while the protein level of collagen II wasincreased by ICA treatment (Fig. 6d, e). Collectively, these

results indicated that ICA exerts a protective effect on OAby regulating pyroptosis reactions.

DiscussionOA is a progressive and deteriorative disorder of articu-lar joint, and current medicines including acetamino-phen, nonsteroidal anti-inflammatory drugs, andduloxetine were only able to relieve the pain and unableto cure the disease. Inflammation and apoptosis are as-sociated with the progress of OA, and anti-inflammatoryagents could prevent OA development [18–21]. ICA, theextract of Epimedium, is known to have anti-oxidative,anti-neuroinflammatory, and anti-apoptotic effects andconsidered to be effective for a range of disorders, suchas neoplasm, Alzheimer’s disease, cerebral ischemia, de-pression, diabetes, and Parkinson’s disease (PD) [22–25].For OA, ICA was identified as an anti-osteoporotic andanti-inflammatory compound through multi-targetmechanisms, including the suppression of NF-κB signal-ing and inhibition of MMP1, MMP3, and MMP13 ex-pression or OPG-RANKL-RANK system via MAPKpathways [13–15]. The present study demonstrated that

Fig. 5 ICA protected rat against osteoarthritis. a H&E staining and safranin O/fast green staining were used to evaluate cartilage histopathology inOA rats, and immunohitochemical analysis was used to detect NLRP3 in OA rat cartilage tissue. b Statistical analysis of NLRP3 positive rate. All theresults were shown as mean ± SD (n = 3), which were three separate experiments performed in triplicate. *p < 0.05 and **p < 0.01

Zu et al. Journal of Orthopaedic Surgery and Research (2019) 14:307 Page 9 of 12

ICA inhibited LPS-mediated chondrocytes injury and pyr-optosis and alleviated rat osteoarthritis by inhibiting NLRP3and caspase-1 signaling. Thus, we uncovered a novel mech-anism by which ICA could prevent OA, further indicatingthe protective effect of ICA in OA treatment.Pathologically, MAPKs/NF-κB-mediated inflammatory

response and cell apoptosis is critically involved in thedevelopment of OA. NLRP3 inflammasome has beensuggested to associate with the pathogenesis of variousarthritic disorders by stimulating proinflammatory cyto-kines and degradative enzymes [26]. A recent studyreported the essential role of NLRP1 and NLRP3 inflam-masomes in the fibroblast-like synoviocytes inflamma-tion and pyroptosis, suggesting NLRP1 and NLPR3inflammasomes could be pathologically important forOA [9]. NLRP3 may serve as a potential biomarker forthe management of OA [26]. Inhibition of NLRP3inflammasome by curcumin or estradiol could downreg-ulate inflammatory cytokines and prevent OA progress[27, 28]. In the present study, ICA attenuated LPS-induced inflammation and pyroptosis by inhibitingNLRP3 inflammasome signaling, and the overexpression

of NLRP3 reduced the protective effect of ICA. Theseresults fully confirmed the pathological effect of NLRP3inflammasome in OA, and NLRP3 inhibition couldachieve a therapeutic effect on OA treatment.Pyroptosis is a caspase-1-dependent type of pro-

grammed cell death activation evoked by inflamma-somes, leading to cellular lysis and cytosolic contentsrelease to the extracellular environment [29]. GasderminD (GSDMD), a member of Gasdermin (GSDM) family,can be cleaved by caspase-1, liberating the N-terminaldomain, which oligomerizes on the plasma membraneforms a pore for the releasing of substrates, such as IL-1β and IL-18 [30]. Caspase-1-mediated pyroptosis wasshown to play a role in the regulation of various dis-eases, such as multiple sclerosis [31], retinitis [32], andneurological disorders[33, 34]. Furthermore, caspase-1-regulated pyroptosis could control Brucella joint infec-tion [35], and the absence of caspase-1 reduced jointpathology in chronic arthritis [36]. The present studyfurther demonstrated that LPS stimulated the expressionof NLRP3, promoted the activity of caspase-1, and in-creased pyroptosis in chondrocytes, highlighting the

Fig. 6 ICA attenuated pyroptosis and promoted collagen formation in OA rats. The mRNA levels of NLRP3 (a), IL-1β (b), and IL-18 (c) wereexamined with qRT-PCR in OA rat cartilage tissue. d The expression level of collagen formation and pyroptosis-related protein was analyzed byWestern blotting in OA rat cartilage tissue. e Quantitative analysis of protein band gray in d. GAPDH was used as internal control. All the resultswere shown as mean ± SD (n = 3), which were three separate experiments performed in triplicate. *p < 0.05 and **p < 0.01

Zu et al. Journal of Orthopaedic Surgery and Research (2019) 14:307 Page 10 of 12

pathological effect of caspase-1-mediated pyroptosis inOA. ICA could interfere with LPS-mediated NLRP3/cas-pase-1 signaling pathway to relief OA.

ConclusionIn conclusion, inflammasome NLRP3 plays a key role inthe pathogenesis of OA. ICA could alleviate pyroptosisby inhibiting NLRP3 signaling-mediated caspase-1 path-way, thereby attenuating the damage of chondrocytesand the occurrence of OA in rats. ICA may be a promis-ing target drug for the treatment of OA

Additional file

Additional files 1: Figure S1. The green fluorescence intensity of thetransfection efficiency increased gradually over time after transfection ofthe NLRP3 overexpression vector. (TIF 1420 kb)

AcknowledgementsNot applicable

Authors’ contributionsYZ conceived and designed the experiments. YM and HJY performed theexperiments. YZ and QL contributed reagents/materials/analysis tools. STZwrote the paper. All authors did the funding acquisition. All authors readand approved the final manuscript.

FundingNot applicable

Availability of data and materialsAll data generated or analyzed during this study are included in thispublished article [and its supplementary information files].

Ethics approval and consent to participateAll procedures were performed according to the internationally acceptedethical guidelines.

Consent for publicationNot applicable

Competing interestsThe authors declare that they have no conflict of interest.

Author details1School of Mechanical and Material Engineering, North China University ofTechnology, No.5, Jinyuanzhuang Road, Shijingshan District, Beijing 100144,People’s Republic of China. 2Department of Stomatology, Peking UnionMedical College Hospital, Chinese Academy of Medical Sciences, Beijing100730, People’s Republic of China.

Received: 11 April 2019 Accepted: 5 August 2019

References1. Guermazi A, et al. Why radiography should no longer be considered a

surrogate outcome measure for longitudinal assessment of cartilage in kneeosteoarthritis. Arthritis Res Ther. 2011;13(6):247.

2. Singh R, Akhtar N, Haqqi TM. Green tea polyphenol epigallocatechin-3-gallate: inflammation and arthritis. [corrected]. Life Sci. 2010;86(25):907–18.

3. Zheng J, Li Q. Methylene blue regulates inflammatory response inosteoarthritis by noncoding long chain RNA CILinc02. J Cell Biochem. 2019;120(3):3331–8.

4. Favero M, et al. Inflammatory molecules produced by meniscus andsynovium in early and end-stage osteoarthritis: a coculture study. J CellPhysiol. 2019;234(7):11176–87.

5. He XF, et al. Investigation for effects of iNOS on biological function ofchondrocytes in rats with post-traumatic osteoarthritis. Eur Rev MedPharmacol Sci. 2018;22(21):7140–7.

6. Patel L, et al. Tenascin-C induces inflammatory mediators and matrixdegradation in osteoarthritic cartilage. BMC Musculoskelet Disord. 2011;12:164.

7. Lu YC, Evans CH, Grodzinsky AJ. Effects of short-term glucocorticoidtreatment on changes in cartilage matrix degradation and chondrocytegene expression induced by mechanical injury and inflammatory cytokines.Arthritis Res Ther. 2011;13(5):R142.

8. Miao EA, Rajan JV, Aderem A. Caspase-1-induced pyroptotic cell death.Immunological Rev. 2011;243(1):206–14.

9. Zhao LR, et al. NLRP1 and NLRP3 inflammasomes mediate LPS/ATP-inducedpyroptosis in knee osteoarthritis. Mol Med Reports. 2018;17:4.

10. Jin C, et al. NLRP3 inflammasome plays a critical role in the pathogenesis ofhydroxyapatite-associated arthropathy. Proc Natl Acad Sci U S A. 2011;108(36):14867–72.

11. Zhou J, et al. Differential miRNAomics of the synovial membrane in kneeosteoarthritis induced by bilateral anterior cruciate ligament transection inrats. Mol Med Rep. 2018;18(4):4051–7.

12. Zhang W, et al. Effect of Chinese traditional herb Epimedium grandiflorumC. Morren and its extract Icariin on osteoarthritis via suppressing NF-kappaBpathway. Indian J Exp Biol. 2013;51(4):313–21.

13. Mi B, et al. Icariin activates autophagy via down-regulation of the NF-κBsignaling-mediated apoptosis in chondrocytes. Front Pharmacol. 2018;9:605.

14. Wang Z, et al. Effects of icariin on the regulation of the OPG-RANKL-RANKsystem are mediated through the MAPK pathways in IL-1β-stimulatedhuman SW1353 chondrosarcoma cells. Int J Mol Med. 2014;34(6):1720–6.

15. Zeng L, et al. Icariin inhibits MMP-1, MMP-3 and MMP-13 expressionthrough MAPK pathways in IL-1β-stimulated SW1353 chondrosarcoma cells.Mol Med Rep. 2017;15(5):2853–8.

16. Vincourt JB, et al. Establishment of a reliable method for direct proteomecharacterization of human articular cartilage. Mol Cell Proteomics. 2006;5(10):1984–95.

17. Parra-Torres NM, Cázares-Raga FE, Kouri JB. Proteomic analysis of ratcartilage: the identification of differentially expressed proteins in the earlystages of osteoarthritis. Proteome Sci. 2014;12(1):55.

18. Qu Y, et al. Morin exhibits anti-inflammatory effects on IL-1β-stimulatedhuman osteoarthritis chondrocytes by activating the Nrf2 signalingpathway. Cell Physiol Biochem. 2018;51(4):1830–8.

19. Scotece M, et al. Oleocanthal inhibits catabolic and inflammatory mediatorsin LPS-activated human primary osteoarthritis (OA) chondrocytes throughMAPKs/NF-κB pathways. Cell Physiol Biochem. 2018;49(6):2414–26.

20. Chin KY, Ima-Nirwana S. The role of vitamin E in preventing and treatingosteoarthritis - a review of the current evidence. Front Pharmacol. 2018;9:946.

21. Zhong G, et al. Artemisinin ameliorates osteoarthritis by inhibiting the Wnt/β-catenin signaling pathway. Cell Physiol Biochem. 2018;51(6):2575–90.

22. Li M, et al. Icariin ameliorates palmitate-induced insulin resistance throughreducing thioredoxin-interacting protein (TXNIP) and suppressing ER stressin C2C12 myotubes. Front Pharmacol. 2018;9:1180.

23. Zhang Y, Kong WN, Chai XQ. Compound of icariin, astragalus, and puerarinmitigates iron overload in the cerebral cortex of Alzheimer’s disease mice.Neural Regen Res. 2018;13(4):731–6.

24. Chen WF, et al. Neuroprotective properties of icariin in MPTP-inducedmouse model of Parkinson’s disease: involvement of PI3K/Akt and MEK/ERKsignaling pathways. Phytomedicine. 2017;25:93–9.

25. Jin J, et al. An outline for the pharmacological effect of icariin in thenervous system. Eur J Pharmacol. 2018;842:20–32.

26. McAllister MJ, et al. NLRP3 as a potentially novel biomarker for themanagement of osteoarthritis. Osteoarthritis Cartilage. 2018;26(5):612–9.

27. Sun Y, et al. Curcumin prevents osteoarthritis by inhibiting the activation ofinflammasome NLRP3. J Interferon Cytokine Res. 2017;37(10):449–55.

28. Shi J, et al. Estradiol inhibits NLRP3 inflammasome in fibroblast-likesynoviocytes activated by lipopolysaccharide and adenosine triphosphate.Int J Rheum Dis. 2017.

29. Man SM, Karki R, Kanneganti TD. Molecular mechanisms and functions ofpyroptosis, inflammatory caspases and inflammasomes in infectiousdiseases. Immunol Rev. 2017;277(1):61-75

30. Feng S, Fox D, Man SM. Mechanisms of Gasdermin family members ininflammasome signaling and cell death. J Mol Biol. 2018;430(18 Pt B):3068–80.

Zu et al. Journal of Orthopaedic Surgery and Research (2019) 14:307 Page 11 of 12

31. McKenzie BA, et al. Caspase-1 inhibition prevents glial inflammasomeactivation and pyroptosis in models of multiple sclerosis. Proc Natl Acad SciU S A. 2018;115(26):E6065–74.

32. Chien H, Dix RD. Evidence for multiple cell death pathways duringdevelopment of experimental cytomegalovirus retinitis in mice withretrovirus-induced immunosuppression: apoptosis, necroptosis, andpyroptosis. J Virol. 2012;86(20):10961–78.

33. Mamik MK, Power C. Inflammasomes in neurological diseases: emergingpathogenic and therapeutic concepts. Brain. 2017;140(9):2273–85.

34. Tan CC, et al. NLRP1 inflammasome is activated in patients with medialtemporal lobe epilepsy and contributes to neuronal pyroptosis in amygdalakindling-induced rat model. J Neuroinflammation. 2015;12:18.

35. Lacey CA, et al. Caspases-1 and caspase-11 mediate pyroptosis,inflammation, and control of. Infect Immun. 2018.

36. Joosten LA, et al. Inflammatory arthritis in caspase 1 gene-deficient mice:contribution of proteinase 3 to caspase 1-independent production ofbioactive interleukin-1beta. Arthritis Rheum. 2009;60(12):3651–62.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Zu et al. Journal of Orthopaedic Surgery and Research (2019) 14:307 Page 12 of 12